Schizophrenia is a devastating illness regardless of the age at which it presents. When this disorder occurs in childhood or adolescence (onset of psychotic symptoms by age 17) the consequences in terms of functional impairment, loss of developmental opportunities, family suffering and societal burden are particularly dramatic. If young patients are refractory to treatment, the long-term outcome is likely to be considerably worse. In recent years, there has been considerable concern that very few data are available on the use of psychotropic drugs in young persons. This is no more apparent than in the antipsychotic treatment arena. Because of the evidence supporting the improved efficacy and/or side effect profile of the second-generation, or atypical antipsychotic medications in adults, and in the case of clozapine, in treatment refractory adults, it is essential to examine whether the potential benefits of these agents can be extended to children and particularly to treatment refractory children. This trial has been specifically targeted for children and adolescents with schizophrenia who have failed separate trials of olanzapine and risperidone at standard doses. Based on existing pilot and other data clozapine and olanzapine are the drugs of choice for this comparison. Specifically, we propose a 12-week, random-assignment, double-blind comparison of clozapine and high-dose olanzapine in 80 children and adolescents (ages 12 to 17) with treatment refractory schizophrenia (onset of psychotic symptoms by age 17). These treatment options represent viable and novel clinical treatments for patients and their families so that all trial participants, regardless of which treatment they are assigned to, will be receiving a new treatment option. Primary outcome measures will include: 1) overall drug response; 2) positive and negative symptoms; 3) adverse effects profiles. This trial will provide a basis to inform clinical treatment decisions and address the critical question as to whether clozapine or high-dose olazapine is more effective in seriously ill children and adolescents who do not respond to other agents.